Counterfactual Reasoning: From Cognition to Computation

This book explores the role of counterfactual reasoning in human reasoning, decision-making, memory, and creativity and reviews the logical, causal, and machine learning formalisms which have been developed to be able to use counterfactuals in intelligent systems. Questions of this sort are fundamental to counterfactual reasoning, the capacity to imagine and understand alternatives and their implications. From a psychological perspective, counterfactual thinking has been acknowledged as a key cognitive process in human cognition for a long time, and it has recently gained significant importance in the fields of artificial intelligence, explainable machine learning, and decision support systems. In this book, researchers and practitioners from across the fields of artificial intelligence, cognitive psychology, and philosophy discuss the diverse aspects of this intriguing concept, from its psychological and philosophical roots to its contemporary computational implementations. It highlights applications of counterfactual explanations, including explainable AI, visual counterfactual generation, and medical image analysis, showing the potential of CFs for increasing transparency, trust, and accountability in high-stakes areas like medicine. The book’s theoretical approach is directly linked to practice, making it useful for researchers, graduate students and AI, ML, computer vision, medical imaging, cognitive science, and related professionals. It gives an introduction as well as future trends in one of the fastest growing fields of explainable and trustworthy AI.